一种改进的基于粗糙集理论的决策树分类算法

A Classification Algorithm Based on Rough Sets

提出一种基于粗糙集理论的决策树分类算法.首先,将核属性集中的核属性进行合取后加入析取变换,实现属性约简;其次,在决策树构造阶段,对各条件属性分别求其上下近似集,进而得到各属性的近似精度.选择近似精度最大的属性作为决策树的根结点,以此方法递归应用到各子树上来选择决策树的结点并实现决策树的剪枝.实例分析表明,改进的算法提高了决策树方法的效率.

The paper proposes a classification method by using the decision tree based on rout sets. First, it reduces the unnecessary attributes when joining the core attributes sets＇ nuclear attribute and transformed by disjunctive; secondly, it gets every condition＇s top and bottom approximation set respectively when constructing the decision tree, and it finally gets every attribute＇s approximation quality. Choosing the max - approximation as the decision tree＇s root nodes, it uses this method through applying it to each sub -tree to choose decision tree＇s nodes and achieves its simplification. The experiment analysis proves that the algorithm proposed in this paper can improve the efficiency of the decision tree method.